This study has two major objectives: 1) to provide a description of the incidence, distribution and pattern of self-reported delinquent behavior across time and to compare the epidemiology of self-reported delinquent behavior with that for officially recorded delinquent behavior and 2) to identify these variables which account for the observed changes in levels and patterns of involvement in delinquent behavior across time. The theoretical conceptualization is based upon an integration of strain and control theory. Two primary paths or sets of conditions which lead youth into involvement in delinquent behavior have been postulated: The first and probably the most frequent sequence involves 1) weak bonds to conventional society and 2) exposure and commitment to groups involved in delinquent activity, while the second involves 1) strong bonds to conventional society, 2) conditions and experiences which serve to attenuate these initial bonds, and 3) exposure and commitment to groups involved in delinquent activity. Basic variables to be measured include integration (occupancy of conventional social roles, relative presence or absence of sanction networks, etc.), commitment (normlessness, attachment to family, personal aspirations/goals, etc.), exposure to deviance, delinquency, and attenuating/reinforcing experiences during adolescence (negative labeling-counter labeling, success-failure experiences). The study design involves a national probability sample of youth. The sample was drawn in 1976 and includes youth aged 11 to 17 in 1976. The sample constitutes a panel which is re-interviewed annually for four successive years. At the descriptive level, age-cohort analyses are planned, along with an analysis of the social correlates of both self-reported and official delinquency. Basic analyses for the identification of developmental sequences and the testing of theoretical propositions will include cluster and typological analyses, multivariate discriminant and regression analyses. Several types of change analysis, utilizing analysis of variance and regression techniques will be used to analyze change in delinquency across time (self-report and official).